1 //
2 // Copyright (c) 2017 The Khronos Group Inc.
3 //
4 // Licensed under the Apache License, Version 2.0 (the "License");
5 // you may not use this file except in compliance with the License.
6 // You may obtain a copy of the License at
7 //
8 // http://www.apache.org/licenses/LICENSE-2.0
9 //
10 // Unless required by applicable law or agreed to in writing, software
11 // distributed under the License is distributed on an "AS IS" BASIS,
12 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 // See the License for the specific language governing permissions and
14 // limitations under the License.
15 //
16 #include "harness/compat.h"
17
18 #include <stdio.h>
19 #include <string.h>
20 #include <sys/types.h>
21 #include <sys/stat.h>
22
23 #include "procs.h"
24
25 static const char *fmax_kernel_code =
26 "__kernel void test_fmax(__global float *srcA, __global float *srcB, __global float *dst)\n"
27 "{\n"
28 " int tid = get_global_id(0);\n"
29 " dst[tid] = fmax(srcA[tid], srcB[tid]);\n"
30 "}\n";
31
32 static const char *fmax2_kernel_code =
33 "__kernel void test_fmax2(__global float2 *srcA, __global float2 *srcB, __global float2 *dst)\n"
34 "{\n"
35 " int tid = get_global_id(0);\n"
36 " dst[tid] = fmax(srcA[tid], srcB[tid]);\n"
37 "}\n";
38
39 static const char *fmax4_kernel_code =
40 "__kernel void test_fmax4(__global float4 *srcA, __global float4 *srcB, __global float4 *dst)\n"
41 "{\n"
42 " int tid = get_global_id(0);\n"
43 " dst[tid] = fmax(srcA[tid], srcB[tid]);\n"
44 "}\n";
45
46 static const char *fmax8_kernel_code =
47 "__kernel void test_fmax8(__global float8 *srcA, __global float8 *srcB, __global float8 *dst)\n"
48 "{\n"
49 " int tid = get_global_id(0);\n"
50 " dst[tid] = fmax(srcA[tid], srcB[tid]);\n"
51 "}\n";
52
53 static const char *fmax16_kernel_code =
54 "__kernel void test_fmax16(__global float16 *srcA, __global float16 *srcB, __global float16 *dst)\n"
55 "{\n"
56 " int tid = get_global_id(0);\n"
57 " dst[tid] = fmax(srcA[tid], srcB[tid]);\n"
58 "}\n";
59
60
61 static const char *fmax3_kernel_code =
62 "__kernel void test_fmax3(__global float *srcA, __global float *srcB, __global float *dst)\n"
63 "{\n"
64 " int tid = get_global_id(0);\n"
65 " vstore3(fmax(vload3(tid,srcA), vload3(tid,srcB)),tid,dst);\n"
66 "}\n";
67
68 static int
verify_fmax(float * inptrA,float * inptrB,float * outptr,int n)69 verify_fmax(float *inptrA, float *inptrB, float *outptr, int n)
70 {
71 float r;
72 int i;
73
74 for (i=0; i<n; i++)
75 {
76 r = (inptrA[i] >= inptrB[i]) ? inptrA[i] : inptrB[i];
77 if (r != outptr[i])
78 return -1;
79 }
80
81 return 0;
82 }
83
84 int
test_fmax(cl_device_id device,cl_context context,cl_command_queue queue,int n_elems)85 test_fmax(cl_device_id device, cl_context context, cl_command_queue queue, int n_elems)
86 {
87 cl_mem streams[3];
88 cl_float *input_ptr[2], *output_ptr, *p;
89 cl_program *program;
90 cl_kernel *kernel;
91 void *values[3];
92 size_t threads[1];
93 int num_elements;
94 int err;
95 int i;
96 MTdata d;
97
98 program = (cl_program*)malloc(sizeof(cl_program)*kTotalVecCount);
99 kernel = (cl_kernel*)malloc(sizeof(cl_kernel)*kTotalVecCount);
100
101 num_elements = n_elems * (1 << (kTotalVecCount-1));
102
103 input_ptr[0] = (cl_float*)malloc(sizeof(cl_float) * num_elements);
104 input_ptr[1] = (cl_float*)malloc(sizeof(cl_float) * num_elements);
105 output_ptr = (cl_float*)malloc(sizeof(cl_float) * num_elements);
106 streams[0] = clCreateBuffer(context, CL_MEM_READ_WRITE,
107 sizeof(cl_float) * num_elements, NULL, NULL);
108 if (!streams[0])
109 {
110 log_error("clCreateBuffer failed\n");
111 return -1;
112 }
113 streams[1] = clCreateBuffer(context, CL_MEM_READ_WRITE,
114 sizeof(cl_float) * num_elements, NULL, NULL);
115 if (!streams[1])
116 {
117 log_error("clCreateBuffer failed\n");
118 return -1;
119 }
120 streams[2] = clCreateBuffer(context, CL_MEM_READ_WRITE,
121 sizeof(cl_float) * num_elements, NULL, NULL);
122 if (!streams[2])
123 {
124 log_error("clCreateBuffer failed\n");
125 return -1;
126 }
127
128 d = init_genrand( gRandomSeed );
129 p = input_ptr[0];
130 for (i=0; i<num_elements; i++)
131 {
132 p[i] = get_random_float(-0x20000000, 0x20000000, d);
133 }
134 p = input_ptr[1];
135 for (i=0; i<num_elements; i++)
136 {
137 p[i] = get_random_float(-0x20000000, 0x20000000,d );
138 }
139 free_mtdata(d); d = NULL;
140
141 err = clEnqueueWriteBuffer( queue, streams[0], true, 0, sizeof(cl_float)*num_elements, (void *)input_ptr[0], 0, NULL, NULL );
142 if (err != CL_SUCCESS)
143 {
144 log_error("clWriteArray failed\n");
145 return -1;
146 }
147 err = clEnqueueWriteBuffer( queue, streams[1], true, 0, sizeof(cl_float)*num_elements, (void *)input_ptr[1], 0, NULL, NULL );
148 if (err != CL_SUCCESS)
149 {
150 log_error("clWriteArray failed\n");
151 return -1;
152 }
153
154 err = create_single_kernel_helper( context, &program[0], &kernel[0], 1, &fmax_kernel_code, "test_fmax" );
155 if (err)
156 return -1;
157 err = create_single_kernel_helper( context, &program[1], &kernel[1], 1, &fmax2_kernel_code, "test_fmax2" );
158 if (err)
159 return -1;
160 err = create_single_kernel_helper( context, &program[2], &kernel[2], 1, &fmax4_kernel_code, "test_fmax4" );
161 if (err)
162 return -1;
163 err = create_single_kernel_helper( context, &program[3], &kernel[3], 1, &fmax8_kernel_code, "test_fmax8" );
164 if (err)
165 return -1;
166 err = create_single_kernel_helper( context, &program[4], &kernel[4], 1, &fmax16_kernel_code, "test_fmax16" );
167 if (err)
168 return -1;
169 err = create_single_kernel_helper( context, &program[5], &kernel[5], 1, &fmax3_kernel_code, "test_fmax3" );
170 if (err)
171 return -1;
172
173
174 values[0] = streams[0];
175 values[1] = streams[1];
176 values[2] = streams[2];
177 for (i=0; i < kTotalVecCount; i++)
178 {
179 err = clSetKernelArg(kernel[i], 0, sizeof streams[0], &streams[0] );
180 err |= clSetKernelArg(kernel[i], 1, sizeof streams[1], &streams[1] );
181 err |= clSetKernelArg(kernel[i], 2, sizeof streams[2], &streams[2] );
182 if (err != CL_SUCCESS)
183 {
184 log_error("clSetKernelArgs failed\n");
185 return -1;
186 }
187 }
188
189 threads[0] = (size_t)n_elems;
190 for (i=0; i < kTotalVecCount; i++)
191 {
192 err = clEnqueueNDRangeKernel( queue, kernel[i], 1, NULL, threads, NULL, 0, NULL, NULL );
193 if (err != CL_SUCCESS)
194 {
195 log_error("clEnqueueNDRangeKernel failed\n");
196 return -1;
197 }
198
199 err = clEnqueueReadBuffer( queue, streams[2], true, 0, sizeof(cl_float)*num_elements, output_ptr, 0, NULL, NULL );
200 if (err != CL_SUCCESS)
201 {
202 log_error("clEnqueueReadBuffer failed\n");
203 return -1;
204 }
205
206 if (verify_fmax(input_ptr[0], input_ptr[1], output_ptr, n_elems*((g_arrVecSizes[i]))))
207 {
208 log_error("FMAX float%d test failed\n", (g_arrVecSizes[i]));
209 err = -1;
210 }
211 else
212 {
213 log_info("FMAX float%d test passed\n", (g_arrVecSizes[i]));
214 err = 0;
215 }
216
217 if (err)
218 break;
219 }
220
221 clReleaseMemObject(streams[0]);
222 clReleaseMemObject(streams[1]);
223 clReleaseMemObject(streams[2]);
224 for (i=0; i < kTotalVecCount; i++)
225 {
226 clReleaseKernel(kernel[i]);
227 clReleaseProgram(program[i]);
228 }
229 free(program);
230 free(kernel);
231 free(input_ptr[0]);
232 free(input_ptr[1]);
233 free(output_ptr);
234
235 return err;
236 }
237
238
239